Software Design-for-Six Sigma (SDFSS) and SEI Technologies meet!

Transcription

1 Pittsburgh, PA Software Design-for-Six Sigma (SDFSS) and SEI Technologies meet! By Robert W. Stoddard Motorola Six Sigma Master Black Belt Senior Member of Technical Staff Software Engineering Institute This material is approved for public release. Sponsored by the U.S. Department of Defense 2005 by Carnegie Mellon University page 1

2 CMU Service Marks The following are service marks of Carnegie Mellon University: Personal Software Process SM PSP SM Team Software Process SM TSP SM The following are registered in the U.S. Patent & Trademark Office by Carnegie Mellon University: Capability Maturity Model CMM Capability Maturity Model Integration CMMI SEI s Architecture Tradeoff Analysis Method ATAM 2005 by Carnegie Mellon University page 2

3 Product Experience 2005 by Carnegie Mellon University page 3

4 SDFSS Opportunities Impacts without SDFSS Benefits with SDFSS Multi-year development projects failed to deliver a working product (min cost=$7m) Failure 1: Did not model performance of new chipset, processor or language Failure 2: Did not adequately characterize the market and business case Failure 3: Did not adequately test the product Business cases modeling all reasonable uncertainties in market and customer segments Schedules with uncertainties modeled Req ts identified, with KJ analysis, to delight customers Design of Experiments used to: optimize and patent fuel efficiency; test object-oriented software; test robustness with fault insertion testing; reduce flight test by 10x 2005 by Carnegie Mellon University page 4

5 Purpose of this Talk To proclaim that software DFSS, within a holistic DFSS approach to product development, is coming of age, To demonstrate that many gaps, in translating traditional DFSS concepts to software engineering, may be solved by the adoption of a number of Software Engineering Institute (SEI) technologies. Thus, DFSS does not have to be re-invented for Software Engineering! 2005 by Carnegie Mellon University page 5

6 Target Audience Executives and Directors contemplating investing in Software DFSS Deployment champions who may be tasked with the training and roll-out of Software DFSS DFSS and Software Engineering Leaders who need to understand both disciplines, and who can lead in translating and interpreting key concepts and tools between the two disciplines by Carnegie Mellon University page 6

7 A Philosophy From the inception of Six Sigma, the overriding objective has been the degree of confidence a customer has that his (or her) product and service-related expectations will be met by the producer. 1 Today, it is a business process that allows companies to drastically improve their bottom line by designing and monitoring everyday business activities in ways that minimize waste and resources while increasing customer satisfaction. 2 1 Michael J. Harry & J. Ronald Lawson, Six Sigma Producibility Analysis and Process Characterization, Addison-Wesley Publishing Company, Inc., Michael J. Harry, & Richard Schroeder, Six Sigma: The Breakthrough Management Strategy Revolutionizing the World s Top Corporations, by Carnegie Mellon University page 7

8 Example Six Sigma Measures The most-cited measure is 3.4 ppm. Other measures defect rate, parts per million (ppm) Sigma level defects per unit (dpu) defects per million opportunities (dpmo) yield 2005 by Carnegie Mellon University page 8

9 DMAIC Alternative Six Sigma Approaches Define-Measure- Analyze-Improve- Control To solve problems and drive incremental business improvements (e.g. fine-tuning an existing process or product) DMAD(O)V Lean Define-Measure- Analyze-Design- (Optimize)-Verify 5S s; Value Stream Mapping; Cycle Time Analysis To solve problems and drive radical improvement (e.g. blowing up and reengineering an existing process or product) To lean a process by simplification, reducing non-value added tasks, optimizing cycle times DFSS Design-For-Six- Sigma Applying the Six Sigma toolkit and intent to product development beginning with portfolio, marketing, engineering, research, sales, and supply chain 2005 by Carnegie Mellon University page 9

10 Holistic View of Software DFSS Portfolio for Six Sigma Marketing for Six Sigma Product Commercialization for Six Sigma Technology Platform R&D for Six Sigma Sales & Distribution for Six Sigma Supply Chain for Six Sigma 2005 by Carnegie Mellon University page 10

11 Holistic View of Software DFSS Portfolio for Six Sigma Using Statistical Methods: Marketing for Six Sigma 1. Identify Opportunities, Markets and Market Segments, Product Commercialization for Six Sigma 2. Gather Long Range Voice of the Customer, 3. Obtain Technology Roadmap and Technology Characterizations from the R&D / Platform Group, Technology Platform R&D for Six Sigma 4. Define Product Portfolio Requirements, 5. Generate Product Portfolio Architectures, Sales & Distribution for Six Sigma 6. Support Portfolio decisions with Real-Options analysis Supply Chain for Six Sigma 7. Evaluate and Select a Product Portfolio, and * Adapted from Clyde Creveling, Marketing for Six Sigma and Product Development with Six Sigma 8. Develop a Prioritized List of Products within a Product Line Strategy by Carnegie Mellon University page 11

12 Holistic View of Software DFSS Portfolio for Six Sigma Marketing for Six Sigma Using Statistical Methods: Product Commercialization for Six Sigma 1. Understand Customer Value and create a Customer Value Dashboard, 2. Develop the Value Stream analysis, Technology Platform R&D for Six Sigma 3. Conducting Marketing FMEA risk analysis and Business SWOT, 4. Creating Marketing Process measures and data collection methods, 5. KJ and Kano Analysis, * Adapted from Clyde Creveling, Marketing for Six Sigma and Product Development with Six Sigma Sales & Distribution for Six Sigma 6. Marketing Composite Design and Product Line Strategies Supply Chain for Six Sigma 7. Market Forecasting, Price Model Planning, Channel Analysis, 8. Portfolio Management, Branding Decision-making and Promotion Analysis 2005 by Carnegie Mellon University page 12

13 Holistic View of Software DFSS Portfolio for Six Sigma Marketing for Six Sigma Product Commercialization for Six Sigma Using Statistical Methods: Technology Platform R&D for Six Sigma 1. Participate in KJ and QFD Voice of the Customer Activities, 2. Develop functional models, architecture and behavioral models, Sales & Distribution for Six Sigma 3. Generate product solution concepts and identify critical parameters for CTQs, Supply Chain for Six Sigma 4. Develop mathematical and statistical models of CTQs, 5. Select solution, implement robust design and track Critical Parameters, 6. Create optimized designs using designed experiments, * Adapted from Clyde Creveling, Marketing for Six Sigma and Product Development with Six Sigma 7. Establish critical parameter tolerances, and verify CTQ achievement by Carnegie Mellon University page 13

14 Holistic View of Software DFSS Portfolio for Six Sigma Product Commercialization for Six Sigma Technology Platform R&D for Six Sigma Using Statistical Methods: Marketing for Six Sigma 1. Characterize existing technologies, capabilities, gaps, risks, expected life Sales & Distribution for Six Sigma 2. Identify and evaluate anticipated technology breakthroughs including probabilistic assessment of timing and capability Supply Chain for Six Sigma 3. Optimize the portfolio of technologies to be pursued in context of the business strategy and latest product portfolio 4. Develop robust platforms that are optimized for the Customer Req ts and CTQs of product lines 5. Develop performance models of platform technology and it s capabilities 2005 by Carnegie Mellon University page 14

15 SEI Technologies Boost DFSS! 2007 European SEPG The SEI Capability Maturity Model Portfolio for Six Sigma Integrated (CMMI) has a product development Marketing for Six Sigma perspective that overlaps Product Commercialization significantly for Six Sigma with all of the DFSS Technology Platform R&D for Six Sigma methodologies! Sales & Distribution for Six Sigma Supply Chain for Six Sigma 2005 by Carnegie Mellon University page 15

16 CMMI & Six Sigma (DFSS) Connections Many connections exist, specifically with the following basic process areas: RD, REQM, TS VER, VAL DAR, RSKM, MA Connections are further highlighted on the following slides with CMMI High Maturity Process Areas: QPM, OPP CAR, OID 2005 by Carnegie Mellon University page 16

17 CMMI OPP and Six Sigma SP1.1 Processes Big Y Business Goal-to-Vital x Process; Processes driving central tendency and variation SP1.2 Measures Critical Parameter Management; CTQ factors; Root Cause Analysis of subprocess factors SP1.3 Objectives SP1.4 Baselines SP1.5 Models KJ Analysis ; Analytic Hierarchy Process; Categorical Survey Data Analysis; Six Sigma Scorecards Control Charts; Graphical Summaries in Minitab; Central Tendency and Variation; Confidence and Prediction Intervals ANOVA; Regression; Chi-Square; Logistic Regression; Monte Carlo Simulation; Discrete Event Process Simulation; Design of Experiments; Response Surface Methodology; Multiple Y Optimization; Probabilistic Models 2005 by Carnegie Mellon University page 17

18 CMMI QPM and Six Sigma SG1 Quantitatively Manage the Project SG2 Statistically Manage Subprocess Performance KJ Analysis ; Analytic Hierarchy Process; Categorical Survey Data Analysis; Six Sigma Scorecards; Big Y Business Goal-to-Vital x Process; Process Mapping Methods and Value- Stream Analysis; Processes driving central tendency and variation; Critical Parameter Management; CTQ factors; Root Cause Analysis of Sub-process factors; Cockpit Control Charts; Graphical Summaries in Minitab; Central Tendency and Variation; Confidence and Prediction Intervals; ANOVA; Regression; Chi- Square; Logistic Regression; Monte Carlo Simulation; Discrete Event Process Simulation; Design of Experiments; Response Surface Methodology; Multiple Y Optimization; Probabilistic Models 2005 by Carnegie Mellon University page 18

19 CMMI CAR and Six Sigma SP1.1 Select Defect Data for Analysis SP1.2 Analyze Causes SP2.1 Implement the Action Proposals SP2.2 Evaluate the Effect of Changes SP2.3 Record Data Measure Phase tools and methods within DMAIC or DMAD(O)V; Models provide insight to the areas of defect data to concentrate on Root Cause Methods, e.g. Ishikawa Diagrams, statistical hypothesis tests to determine if segments are different Piloting; Comparative Studies; Technological and Cultural Change Management techniques Before and After studies and Hypothesis tests; Survey categorical data analysis; compare to results of prediction models Study results; Lessons Learned shared across the organization; Institutional learning 2005 by Carnegie Mellon University page 19

20 CMMI OID and Six Sigma SG1 Select Improvements Six Sigma Big Y to Vital x semi-annual workshops; Business Goal simulation and optimization models; Benchmarking; Capability data sharing; Theory of Inventing (TRIZ) methods; Usage of performance models to identify the major opportunities for improvement with innovation; Assumption Busters; Empowered innovative thinking; Incentives for Innovation; Strong Teaming for Innovation; Various decision models such as AHP, Pugh Method, Probabilistic decision trees SG2 Deploy Improvements Process and Design FMEA; Organizational Readiness for Change; Change Agents; Sponsors; Champions; Influence Leaders; Adoption Curve; Piloting; Risk-based deployment; Before and After comparisons with Hypothesis tests; Results compared to prediction models; Proactive mitigation of risks 2005 by Carnegie Mellon University page 20

21 The CMMI, TSP and PSP Reference: The Great Myths About TSP SM and CMMI by McHale, by Carnegie Mellon University page 21

22 SEI Technologies Boost DFSS! Portfolio for Six Sigma Marketing for Six Sigma Product Commercialization for Six Sigma Technology Platform R&D for Six Sigma The SEI Team Sales Software & Distribution Process for (TSP) Six and Sigma Personal Software Process (PSP) significantly enhance Supply the Chain software for Six development Sigma teaming within Product Commercialization and Technology Platform R&D for Six Sigma! 2005 by Carnegie Mellon University page 22

23 The Personal Software Process 2007 European SEPG Reference: The TSP and PSP Tutorial by the SEI (Burton, Davis, McHale), by Carnegie Mellon University page 23

24 TSP Builds Software Teams Reference: The TSP and PSP Tutorial by the SEI (Burton, Davis, McHale), by Carnegie Mellon University page 24

25 The TSP Launch Process Reference: The TSP and PSP Tutorial by the SEI (Burton, Davis, McHale), by Carnegie Mellon University page 25

26 SEI Technologies Boost DFSS! The SEI Product Line Portfolio for Six Sigma Practice work overlaps significantly with both Portfolio for Six Sigma, and Marketing for Six Sigma Technology Platform R&D for Six Sigma! Product Commercialization for Six Sigma Technology Platform R&D for Six Sigma Sales & Distribution for Six Sigma Supply Chain for Six Sigma 2005 by Carnegie Mellon University page 26

27 Reference: Software Product Lines, Paul Clements, Linda Northrop, by Carnegie Mellon University page 27

28 Product Line Practice Framework Reference: Software Product Lines, Paul Clements, Linda Northrop, by Carnegie Mellon University page 28

29 SEI Technologies Boost DFSS! Portfolio for Six Sigma Marketing for Six Sigma Product Commercialization for Six Sigma Technology Platform R&D for Six Sigma The SEI Sales Software & Distribution Architecture for Six work Sigma overlaps significantly with both Portfolio for Supply Six Sigma, Chain Product for Six Commercialization Sigma for Six Sigma, and Technology Platform R&D for Six Sigma! 2005 by Carnegie Mellon University page 29

30 Reference: Future Directions of the Software Architecture Technology Initiative, Second Annual SATURN Workshop, Mark Klein, by Carnegie Mellon University page 30

31 Reference: Future Directions of the Software Architecture Technology Initiative, Second Annual SATURN Workshop, Mark Klein, by Carnegie Mellon University page 31

32 Reference: Future Directions of the Software Architecture Technology Initiative, SATURN Workshop, Mark Klein, by Carnegie Mellon University page 32

33 Reference: Future Directions of the Software Architecture Technology Initiative, Second Annual SATURN Workshop, Mark Klein, by Carnegie Mellon University page 33

34 In Summary Software DFSS, within a holistic DFSS approach to product development, is coming of age, Many gaps, in translating traditional DFSS concepts to software engineering, may be solved by the adoption of a number of Software Engineering Institute (SEI) technologies. Thus, DFSS does not have to be re-invented for Software Engineering! 2005 by Carnegie Mellon University page 34

35 Contact Information Robert Stoddard Software Engineering Institute (Office 3110) 4500 Fifth Avenue Pittsburgh, PA Or, contact SEI Customer Relations: Customer Relations Software Engineering Institute Carnegie Mellon University Pittsburgh, PA FAX: (412) by Carnegie Mellon University page 35

36 Questions 2005 by Carnegie Mellon University page 36